四、(2025·黑龙江大庆三模)完形填空。
DeepSeek-R1 is
B
AI model from China. It can
C
out maths problems, write simple codes(代码), and talk like people in daily life. Different from other AI models, it learns by doing lots of tasks every day, not just reading examples. This makes it very special.
Most AI models need many
A
from books to learn, but R1 learns by itself. With more practice, it becomes
C
. Engineers say, “We let it try again and again, just like students do exercises. Soon, it can finish tasks well.”
What is the best thing? Schools and small shops can use it for
A
. A village teacher says, “We don’t worry about
C
now. Even poor places can get help from this smart AI tool. It changes our daily work.”
But R1 still has problems. It makes mistakes in languages
C
German. Engineers try to
B
these mistakes and correct them. They say, “We want R1 to help people
D
the world.”
This AI tool teaches us an important lesson—great things need to
B
more. Just having money cannot make something amazing. Let’s learn from R1!
1. A. a B. an C. the D. /
2. A. play B. get C. work D. find
3. A. examples B. games C. pictures D. questions
4. A. good B. interesting C. smart D. different
5. A. work B. exercise C. help D. advice
6. A. news B. time C. money D. books
7. A. with B. from C. like D. for
8. A. forget B. mark C. answer D. remember
9. A. a B. of C. on D. around
10. A. love B. try C. help D. see